Ontology based system to integrate phenotype data from multiple studies

The rat has long been used in biomedical research and is considered the premier organism for disease and physiological studies. Numerous strains with particular phenotypes have been developed to serve as models of human disease. The first complete rat genome for the BN/NHsdMcwi has been followed by several projects to completely sequence additional strains, including those which serve as disease models. The availability of the genome sequence for multiple strains of rat now makes it possible for investigators to conduct more precise studies on the genetic basis of disease and the relationship of the genotype to phenotype. At the heart of such studies is the accurate collection, integration, and warehousing of the comprehensive phenotype data required to conduct Genome Wide Association Studies or other statistical methods to detect and quantify the association between genotype and phenotype. The objective of this work was to create an ontology based system for standardizing, integrating and presenting rat phenotype measurement data from multiple sources. In order to accomplish this objective, the PhenoMiner project was launched. The PhenoMiner project consists of three ontologies for standardizing elements of phenotype measurement records, a database, a data mining and visualization tool, and a data curation tool. The three ontologies were created to facilitate the standardization of phenotype measurement data that is necessary for integration. The Clinical Measurement Ontology, Measurement Method Ontology and Experimental Condition Ontology were developed to standardize data about 1) what was measured, 2) how it was measured and 3) under what conditions it was measured. The PhenoMiner relational database was constructed to house data from multiple sources. Using this database, data from two major phenotyping projects and the literature were integrated. The PhenoMiner data mining and visualization tool allows users to search by rat strain, clinical measurement type, measurement method type or experimental condition. Multiple filters and tracking systems allow the user to customize queries for optimal results. These results are visualized in a bar chart that allows further filtering and customization and data can be downloaded in several formats. PhenoMiner provides the most comprehensive resource for creating individual strain profiles, for comparing phenotypes across strains, for the identification of the effect of sex, measurement method and experimental conditions on measurement values.